{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.5845389449072045,
  "eval_steps": 500,
  "global_step": 1000,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.001169077889814409,
      "grad_norm": 6.78125,
      "learning_rate": 3.898635477582846e-07,
      "loss": 5.0851,
      "step": 2
    },
    {
      "epoch": 0.002338155779628818,
      "grad_norm": 7.0625,
      "learning_rate": 7.797270955165692e-07,
      "loss": 5.1921,
      "step": 4
    },
    {
      "epoch": 0.0035072336694432268,
      "grad_norm": 7.34375,
      "learning_rate": 1.1695906432748538e-06,
      "loss": 5.0756,
      "step": 6
    },
    {
      "epoch": 0.004676311559257636,
      "grad_norm": 7.03125,
      "learning_rate": 1.5594541910331385e-06,
      "loss": 5.1097,
      "step": 8
    },
    {
      "epoch": 0.005845389449072045,
      "grad_norm": 6.40625,
      "learning_rate": 1.949317738791423e-06,
      "loss": 5.0591,
      "step": 10
    },
    {
      "epoch": 0.0070144673388864535,
      "grad_norm": 5.8125,
      "learning_rate": 2.3391812865497075e-06,
      "loss": 5.096,
      "step": 12
    },
    {
      "epoch": 0.008183545228700862,
      "grad_norm": 5.1875,
      "learning_rate": 2.729044834307992e-06,
      "loss": 5.0234,
      "step": 14
    },
    {
      "epoch": 0.009352623118515271,
      "grad_norm": 5.03125,
      "learning_rate": 3.118908382066277e-06,
      "loss": 5.0266,
      "step": 16
    },
    {
      "epoch": 0.01052170100832968,
      "grad_norm": 4.90625,
      "learning_rate": 3.5087719298245615e-06,
      "loss": 5.031,
      "step": 18
    },
    {
      "epoch": 0.01169077889814409,
      "grad_norm": 4.90625,
      "learning_rate": 3.898635477582846e-06,
      "loss": 4.9845,
      "step": 20
    },
    {
      "epoch": 0.012859856787958498,
      "grad_norm": 4.6875,
      "learning_rate": 4.2884990253411305e-06,
      "loss": 4.9933,
      "step": 22
    },
    {
      "epoch": 0.014028934677772907,
      "grad_norm": 4.625,
      "learning_rate": 4.678362573099415e-06,
      "loss": 5.0822,
      "step": 24
    },
    {
      "epoch": 0.015198012567587316,
      "grad_norm": 4.71875,
      "learning_rate": 5.0682261208576995e-06,
      "loss": 4.9392,
      "step": 26
    },
    {
      "epoch": 0.016367090457401725,
      "grad_norm": 4.65625,
      "learning_rate": 5.458089668615984e-06,
      "loss": 5.0441,
      "step": 28
    },
    {
      "epoch": 0.017536168347216132,
      "grad_norm": 4.6875,
      "learning_rate": 5.8479532163742686e-06,
      "loss": 4.8865,
      "step": 30
    },
    {
      "epoch": 0.018705246237030543,
      "grad_norm": 4.59375,
      "learning_rate": 6.237816764132554e-06,
      "loss": 4.9227,
      "step": 32
    },
    {
      "epoch": 0.01987432412684495,
      "grad_norm": 4.65625,
      "learning_rate": 6.6276803118908384e-06,
      "loss": 4.9035,
      "step": 34
    },
    {
      "epoch": 0.02104340201665936,
      "grad_norm": 4.59375,
      "learning_rate": 7.017543859649123e-06,
      "loss": 4.863,
      "step": 36
    },
    {
      "epoch": 0.022212479906473768,
      "grad_norm": 4.71875,
      "learning_rate": 7.4074074074074075e-06,
      "loss": 4.7735,
      "step": 38
    },
    {
      "epoch": 0.02338155779628818,
      "grad_norm": 4.84375,
      "learning_rate": 7.797270955165692e-06,
      "loss": 4.7932,
      "step": 40
    },
    {
      "epoch": 0.024550635686102586,
      "grad_norm": 5.3125,
      "learning_rate": 8.187134502923977e-06,
      "loss": 4.6933,
      "step": 42
    },
    {
      "epoch": 0.025719713575916996,
      "grad_norm": 7.15625,
      "learning_rate": 8.576998050682261e-06,
      "loss": 4.5906,
      "step": 44
    },
    {
      "epoch": 0.026888791465731404,
      "grad_norm": 8.8125,
      "learning_rate": 8.966861598440546e-06,
      "loss": 4.4291,
      "step": 46
    },
    {
      "epoch": 0.028057869355545814,
      "grad_norm": 9.6875,
      "learning_rate": 9.35672514619883e-06,
      "loss": 4.221,
      "step": 48
    },
    {
      "epoch": 0.02922694724536022,
      "grad_norm": 8.8125,
      "learning_rate": 9.746588693957115e-06,
      "loss": 4.032,
      "step": 50
    },
    {
      "epoch": 0.030396025135174632,
      "grad_norm": 9.0,
      "learning_rate": 1.0136452241715399e-05,
      "loss": 3.8603,
      "step": 52
    },
    {
      "epoch": 0.03156510302498904,
      "grad_norm": 7.46875,
      "learning_rate": 1.0526315789473684e-05,
      "loss": 3.6522,
      "step": 54
    },
    {
      "epoch": 0.03273418091480345,
      "grad_norm": 8.375,
      "learning_rate": 1.0916179337231968e-05,
      "loss": 3.4394,
      "step": 56
    },
    {
      "epoch": 0.03390325880461786,
      "grad_norm": 6.4375,
      "learning_rate": 1.1306042884990253e-05,
      "loss": 3.2265,
      "step": 58
    },
    {
      "epoch": 0.035072336694432264,
      "grad_norm": 5.75,
      "learning_rate": 1.1695906432748537e-05,
      "loss": 3.091,
      "step": 60
    },
    {
      "epoch": 0.03624141458424668,
      "grad_norm": 4.8125,
      "learning_rate": 1.2085769980506823e-05,
      "loss": 2.8856,
      "step": 62
    },
    {
      "epoch": 0.037410492474061086,
      "grad_norm": 3.8125,
      "learning_rate": 1.2475633528265108e-05,
      "loss": 2.6299,
      "step": 64
    },
    {
      "epoch": 0.03857957036387549,
      "grad_norm": 3.234375,
      "learning_rate": 1.2865497076023392e-05,
      "loss": 2.5917,
      "step": 66
    },
    {
      "epoch": 0.0397486482536899,
      "grad_norm": 4.0625,
      "learning_rate": 1.3255360623781677e-05,
      "loss": 2.3757,
      "step": 68
    },
    {
      "epoch": 0.040917726143504314,
      "grad_norm": 2.453125,
      "learning_rate": 1.364522417153996e-05,
      "loss": 2.308,
      "step": 70
    },
    {
      "epoch": 0.04208680403331872,
      "grad_norm": 2.390625,
      "learning_rate": 1.4035087719298246e-05,
      "loss": 2.3042,
      "step": 72
    },
    {
      "epoch": 0.04325588192313313,
      "grad_norm": 2.109375,
      "learning_rate": 1.442495126705653e-05,
      "loss": 2.2149,
      "step": 74
    },
    {
      "epoch": 0.044424959812947536,
      "grad_norm": 2.71875,
      "learning_rate": 1.4814814814814815e-05,
      "loss": 2.0738,
      "step": 76
    },
    {
      "epoch": 0.04559403770276195,
      "grad_norm": 2.46875,
      "learning_rate": 1.5204678362573099e-05,
      "loss": 1.9663,
      "step": 78
    },
    {
      "epoch": 0.04676311559257636,
      "grad_norm": 2.640625,
      "learning_rate": 1.5594541910331384e-05,
      "loss": 1.9384,
      "step": 80
    },
    {
      "epoch": 0.047932193482390764,
      "grad_norm": 2.0,
      "learning_rate": 1.5984405458089668e-05,
      "loss": 1.894,
      "step": 82
    },
    {
      "epoch": 0.04910127137220517,
      "grad_norm": 2.078125,
      "learning_rate": 1.6374269005847955e-05,
      "loss": 1.8204,
      "step": 84
    },
    {
      "epoch": 0.050270349262019585,
      "grad_norm": 2.953125,
      "learning_rate": 1.676413255360624e-05,
      "loss": 1.8051,
      "step": 86
    },
    {
      "epoch": 0.05143942715183399,
      "grad_norm": 2.046875,
      "learning_rate": 1.7153996101364522e-05,
      "loss": 1.5832,
      "step": 88
    },
    {
      "epoch": 0.0526085050416484,
      "grad_norm": 1.75,
      "learning_rate": 1.7543859649122806e-05,
      "loss": 1.5298,
      "step": 90
    },
    {
      "epoch": 0.05377758293146281,
      "grad_norm": 1.71875,
      "learning_rate": 1.7933723196881093e-05,
      "loss": 1.4964,
      "step": 92
    },
    {
      "epoch": 0.054946660821277214,
      "grad_norm": 1.3515625,
      "learning_rate": 1.8323586744639376e-05,
      "loss": 1.3987,
      "step": 94
    },
    {
      "epoch": 0.05611573871109163,
      "grad_norm": 1.9140625,
      "learning_rate": 1.871345029239766e-05,
      "loss": 1.3959,
      "step": 96
    },
    {
      "epoch": 0.057284816600906036,
      "grad_norm": 1.34375,
      "learning_rate": 1.9103313840155944e-05,
      "loss": 1.3673,
      "step": 98
    },
    {
      "epoch": 0.05845389449072044,
      "grad_norm": 1.09375,
      "learning_rate": 1.949317738791423e-05,
      "loss": 1.2864,
      "step": 100
    },
    {
      "epoch": 0.05962297238053485,
      "grad_norm": 1.1328125,
      "learning_rate": 1.9883040935672515e-05,
      "loss": 1.2535,
      "step": 102
    },
    {
      "epoch": 0.060792050270349264,
      "grad_norm": 1.2109375,
      "learning_rate": 2.0272904483430798e-05,
      "loss": 1.2242,
      "step": 104
    },
    {
      "epoch": 0.06196112816016367,
      "grad_norm": 2.59375,
      "learning_rate": 2.0662768031189085e-05,
      "loss": 1.1936,
      "step": 106
    },
    {
      "epoch": 0.06313020604997809,
      "grad_norm": 1.953125,
      "learning_rate": 2.105263157894737e-05,
      "loss": 1.1858,
      "step": 108
    },
    {
      "epoch": 0.06429928393979249,
      "grad_norm": 2.140625,
      "learning_rate": 2.1442495126705653e-05,
      "loss": 1.1438,
      "step": 110
    },
    {
      "epoch": 0.0654683618296069,
      "grad_norm": 1.09375,
      "learning_rate": 2.1832358674463936e-05,
      "loss": 1.1217,
      "step": 112
    },
    {
      "epoch": 0.06663743971942131,
      "grad_norm": 1.8671875,
      "learning_rate": 2.2222222222222223e-05,
      "loss": 1.1508,
      "step": 114
    },
    {
      "epoch": 0.06780651760923571,
      "grad_norm": 1.046875,
      "learning_rate": 2.2612085769980507e-05,
      "loss": 1.0976,
      "step": 116
    },
    {
      "epoch": 0.06897559549905012,
      "grad_norm": 1.234375,
      "learning_rate": 2.300194931773879e-05,
      "loss": 1.0716,
      "step": 118
    },
    {
      "epoch": 0.07014467338886453,
      "grad_norm": 1.6171875,
      "learning_rate": 2.3391812865497074e-05,
      "loss": 1.0152,
      "step": 120
    },
    {
      "epoch": 0.07131375127867894,
      "grad_norm": 3.125,
      "learning_rate": 2.378167641325536e-05,
      "loss": 1.0265,
      "step": 122
    },
    {
      "epoch": 0.07248282916849336,
      "grad_norm": 3.546875,
      "learning_rate": 2.4171539961013645e-05,
      "loss": 1.2355,
      "step": 124
    },
    {
      "epoch": 0.07365190705830776,
      "grad_norm": 1.1796875,
      "learning_rate": 2.456140350877193e-05,
      "loss": 1.0111,
      "step": 126
    },
    {
      "epoch": 0.07482098494812217,
      "grad_norm": 2.453125,
      "learning_rate": 2.4951267056530216e-05,
      "loss": 1.2317,
      "step": 128
    },
    {
      "epoch": 0.07599006283793658,
      "grad_norm": 1.09375,
      "learning_rate": 2.53411306042885e-05,
      "loss": 0.9988,
      "step": 130
    },
    {
      "epoch": 0.07715914072775099,
      "grad_norm": 1.09375,
      "learning_rate": 2.5730994152046783e-05,
      "loss": 1.0076,
      "step": 132
    },
    {
      "epoch": 0.07832821861756539,
      "grad_norm": 1.4375,
      "learning_rate": 2.6120857699805067e-05,
      "loss": 1.0155,
      "step": 134
    },
    {
      "epoch": 0.0794972965073798,
      "grad_norm": 1.796875,
      "learning_rate": 2.6510721247563354e-05,
      "loss": 1.1216,
      "step": 136
    },
    {
      "epoch": 0.08066637439719421,
      "grad_norm": 0.9921875,
      "learning_rate": 2.6900584795321637e-05,
      "loss": 0.9788,
      "step": 138
    },
    {
      "epoch": 0.08183545228700863,
      "grad_norm": 1.328125,
      "learning_rate": 2.729044834307992e-05,
      "loss": 0.9372,
      "step": 140
    },
    {
      "epoch": 0.08300453017682304,
      "grad_norm": 1.296875,
      "learning_rate": 2.7680311890838205e-05,
      "loss": 0.897,
      "step": 142
    },
    {
      "epoch": 0.08417360806663744,
      "grad_norm": 1.0703125,
      "learning_rate": 2.8070175438596492e-05,
      "loss": 0.9766,
      "step": 144
    },
    {
      "epoch": 0.08534268595645185,
      "grad_norm": 0.84765625,
      "learning_rate": 2.8460038986354776e-05,
      "loss": 0.9315,
      "step": 146
    },
    {
      "epoch": 0.08651176384626626,
      "grad_norm": 0.9140625,
      "learning_rate": 2.884990253411306e-05,
      "loss": 0.9425,
      "step": 148
    },
    {
      "epoch": 0.08768084173608066,
      "grad_norm": 0.99609375,
      "learning_rate": 2.9239766081871346e-05,
      "loss": 0.9242,
      "step": 150
    },
    {
      "epoch": 0.08884991962589507,
      "grad_norm": 1.8046875,
      "learning_rate": 2.962962962962963e-05,
      "loss": 0.9396,
      "step": 152
    },
    {
      "epoch": 0.09001899751570948,
      "grad_norm": 0.92578125,
      "learning_rate": 3.0019493177387914e-05,
      "loss": 0.8932,
      "step": 154
    },
    {
      "epoch": 0.0911880754055239,
      "grad_norm": 72.5,
      "learning_rate": 3.0409356725146197e-05,
      "loss": 0.9062,
      "step": 156
    },
    {
      "epoch": 0.0923571532953383,
      "grad_norm": 1.703125,
      "learning_rate": 3.0799220272904484e-05,
      "loss": 0.9534,
      "step": 158
    },
    {
      "epoch": 0.09352623118515271,
      "grad_norm": 1.453125,
      "learning_rate": 3.118908382066277e-05,
      "loss": 0.8715,
      "step": 160
    },
    {
      "epoch": 0.09469530907496712,
      "grad_norm": 1.015625,
      "learning_rate": 3.157894736842105e-05,
      "loss": 0.9029,
      "step": 162
    },
    {
      "epoch": 0.09586438696478153,
      "grad_norm": 1.15625,
      "learning_rate": 3.1968810916179335e-05,
      "loss": 0.8968,
      "step": 164
    },
    {
      "epoch": 0.09703346485459594,
      "grad_norm": 1.4296875,
      "learning_rate": 3.235867446393762e-05,
      "loss": 0.9784,
      "step": 166
    },
    {
      "epoch": 0.09820254274441034,
      "grad_norm": 1.0078125,
      "learning_rate": 3.274853801169591e-05,
      "loss": 0.9012,
      "step": 168
    },
    {
      "epoch": 0.09937162063422475,
      "grad_norm": 0.8359375,
      "learning_rate": 3.313840155945419e-05,
      "loss": 0.8895,
      "step": 170
    },
    {
      "epoch": 0.10054069852403917,
      "grad_norm": 0.8046875,
      "learning_rate": 3.352826510721248e-05,
      "loss": 0.8858,
      "step": 172
    },
    {
      "epoch": 0.10170977641385358,
      "grad_norm": 0.83203125,
      "learning_rate": 3.391812865497076e-05,
      "loss": 0.8408,
      "step": 174
    },
    {
      "epoch": 0.10287885430366799,
      "grad_norm": 0.8671875,
      "learning_rate": 3.4307992202729044e-05,
      "loss": 0.8896,
      "step": 176
    },
    {
      "epoch": 0.10404793219348239,
      "grad_norm": 1.78125,
      "learning_rate": 3.469785575048733e-05,
      "loss": 0.8941,
      "step": 178
    },
    {
      "epoch": 0.1052170100832968,
      "grad_norm": 0.8984375,
      "learning_rate": 3.508771929824561e-05,
      "loss": 0.9013,
      "step": 180
    },
    {
      "epoch": 0.1063860879731112,
      "grad_norm": 1.2734375,
      "learning_rate": 3.5477582846003895e-05,
      "loss": 0.8585,
      "step": 182
    },
    {
      "epoch": 0.10755516586292561,
      "grad_norm": 2.015625,
      "learning_rate": 3.5867446393762186e-05,
      "loss": 1.0231,
      "step": 184
    },
    {
      "epoch": 0.10872424375274002,
      "grad_norm": 1.7578125,
      "learning_rate": 3.625730994152047e-05,
      "loss": 0.8835,
      "step": 186
    },
    {
      "epoch": 0.10989332164255443,
      "grad_norm": 1.078125,
      "learning_rate": 3.664717348927875e-05,
      "loss": 0.8465,
      "step": 188
    },
    {
      "epoch": 0.11106239953236885,
      "grad_norm": 2.75,
      "learning_rate": 3.7037037037037037e-05,
      "loss": 1.1014,
      "step": 190
    },
    {
      "epoch": 0.11223147742218326,
      "grad_norm": 2.09375,
      "learning_rate": 3.742690058479532e-05,
      "loss": 0.8691,
      "step": 192
    },
    {
      "epoch": 0.11340055531199766,
      "grad_norm": 0.96875,
      "learning_rate": 3.7816764132553604e-05,
      "loss": 0.8547,
      "step": 194
    },
    {
      "epoch": 0.11456963320181207,
      "grad_norm": 1.5859375,
      "learning_rate": 3.820662768031189e-05,
      "loss": 0.8725,
      "step": 196
    },
    {
      "epoch": 0.11573871109162648,
      "grad_norm": 0.88671875,
      "learning_rate": 3.859649122807018e-05,
      "loss": 0.8393,
      "step": 198
    },
    {
      "epoch": 0.11690778898144089,
      "grad_norm": 1.015625,
      "learning_rate": 3.898635477582846e-05,
      "loss": 0.8345,
      "step": 200
    },
    {
      "epoch": 0.11807686687125529,
      "grad_norm": 0.82421875,
      "learning_rate": 3.9376218323586745e-05,
      "loss": 0.8966,
      "step": 202
    },
    {
      "epoch": 0.1192459447610697,
      "grad_norm": 0.79296875,
      "learning_rate": 3.976608187134503e-05,
      "loss": 0.8291,
      "step": 204
    },
    {
      "epoch": 0.12041502265088412,
      "grad_norm": 0.83203125,
      "learning_rate": 4.015594541910331e-05,
      "loss": 0.8175,
      "step": 206
    },
    {
      "epoch": 0.12158410054069853,
      "grad_norm": 0.74609375,
      "learning_rate": 4.0545808966861596e-05,
      "loss": 0.8353,
      "step": 208
    },
    {
      "epoch": 0.12275317843051294,
      "grad_norm": 1.796875,
      "learning_rate": 4.093567251461988e-05,
      "loss": 1.0084,
      "step": 210
    },
    {
      "epoch": 0.12392225632032734,
      "grad_norm": 59.75,
      "learning_rate": 4.132553606237817e-05,
      "loss": 0.8478,
      "step": 212
    },
    {
      "epoch": 0.12509133421014176,
      "grad_norm": 1.265625,
      "learning_rate": 4.1715399610136454e-05,
      "loss": 0.8267,
      "step": 214
    },
    {
      "epoch": 0.12626041209995617,
      "grad_norm": 1.0625,
      "learning_rate": 4.210526315789474e-05,
      "loss": 0.836,
      "step": 216
    },
    {
      "epoch": 0.12742948998977058,
      "grad_norm": 1.0078125,
      "learning_rate": 4.249512670565302e-05,
      "loss": 0.857,
      "step": 218
    },
    {
      "epoch": 0.12859856787958499,
      "grad_norm": 0.92578125,
      "learning_rate": 4.2884990253411305e-05,
      "loss": 0.8478,
      "step": 220
    },
    {
      "epoch": 0.1297676457693994,
      "grad_norm": 0.953125,
      "learning_rate": 4.327485380116959e-05,
      "loss": 0.844,
      "step": 222
    },
    {
      "epoch": 0.1309367236592138,
      "grad_norm": 0.796875,
      "learning_rate": 4.366471734892787e-05,
      "loss": 0.8128,
      "step": 224
    },
    {
      "epoch": 0.1321058015490282,
      "grad_norm": 1.28125,
      "learning_rate": 4.4054580896686156e-05,
      "loss": 0.8276,
      "step": 226
    },
    {
      "epoch": 0.13327487943884261,
      "grad_norm": 0.94921875,
      "learning_rate": 4.4444444444444447e-05,
      "loss": 0.8478,
      "step": 228
    },
    {
      "epoch": 0.13444395732865702,
      "grad_norm": 1.0703125,
      "learning_rate": 4.483430799220273e-05,
      "loss": 0.814,
      "step": 230
    },
    {
      "epoch": 0.13561303521847143,
      "grad_norm": 1.1328125,
      "learning_rate": 4.5224171539961014e-05,
      "loss": 0.765,
      "step": 232
    },
    {
      "epoch": 0.13678211310828584,
      "grad_norm": 0.87109375,
      "learning_rate": 4.56140350877193e-05,
      "loss": 0.8277,
      "step": 234
    },
    {
      "epoch": 0.13795119099810024,
      "grad_norm": 0.921875,
      "learning_rate": 4.600389863547758e-05,
      "loss": 0.8103,
      "step": 236
    },
    {
      "epoch": 0.13912026888791465,
      "grad_norm": 1.4921875,
      "learning_rate": 4.6393762183235865e-05,
      "loss": 0.7897,
      "step": 238
    },
    {
      "epoch": 0.14028934677772906,
      "grad_norm": 0.7734375,
      "learning_rate": 4.678362573099415e-05,
      "loss": 0.7798,
      "step": 240
    },
    {
      "epoch": 0.14145842466754346,
      "grad_norm": 0.98828125,
      "learning_rate": 4.717348927875244e-05,
      "loss": 0.8442,
      "step": 242
    },
    {
      "epoch": 0.14262750255735787,
      "grad_norm": 0.91796875,
      "learning_rate": 4.756335282651072e-05,
      "loss": 0.7936,
      "step": 244
    },
    {
      "epoch": 0.1437965804471723,
      "grad_norm": 0.859375,
      "learning_rate": 4.7953216374269006e-05,
      "loss": 0.7726,
      "step": 246
    },
    {
      "epoch": 0.1449656583369867,
      "grad_norm": 1.3671875,
      "learning_rate": 4.834307992202729e-05,
      "loss": 0.9626,
      "step": 248
    },
    {
      "epoch": 0.14613473622680112,
      "grad_norm": 0.921875,
      "learning_rate": 4.8732943469785574e-05,
      "loss": 0.7929,
      "step": 250
    },
    {
      "epoch": 0.14730381411661553,
      "grad_norm": 0.92578125,
      "learning_rate": 4.912280701754386e-05,
      "loss": 0.8589,
      "step": 252
    },
    {
      "epoch": 0.14847289200642994,
      "grad_norm": 0.83203125,
      "learning_rate": 4.951267056530214e-05,
      "loss": 0.7383,
      "step": 254
    },
    {
      "epoch": 0.14964196989624434,
      "grad_norm": 1.046875,
      "learning_rate": 4.990253411306043e-05,
      "loss": 0.8011,
      "step": 256
    },
    {
      "epoch": 0.15081104778605875,
      "grad_norm": 1.3359375,
      "learning_rate": 5.0292397660818715e-05,
      "loss": 0.8404,
      "step": 258
    },
    {
      "epoch": 0.15198012567587316,
      "grad_norm": 1.21875,
      "learning_rate": 5.0682261208577e-05,
      "loss": 0.8301,
      "step": 260
    },
    {
      "epoch": 0.15314920356568756,
      "grad_norm": 0.93359375,
      "learning_rate": 5.107212475633528e-05,
      "loss": 0.8157,
      "step": 262
    },
    {
      "epoch": 0.15431828145550197,
      "grad_norm": 1.3203125,
      "learning_rate": 5.1461988304093566e-05,
      "loss": 0.7724,
      "step": 264
    },
    {
      "epoch": 0.15548735934531638,
      "grad_norm": 1.09375,
      "learning_rate": 5.185185185185185e-05,
      "loss": 0.7586,
      "step": 266
    },
    {
      "epoch": 0.15665643723513079,
      "grad_norm": 1.1484375,
      "learning_rate": 5.2241715399610133e-05,
      "loss": 0.7812,
      "step": 268
    },
    {
      "epoch": 0.1578255151249452,
      "grad_norm": 0.74609375,
      "learning_rate": 5.2631578947368424e-05,
      "loss": 0.7486,
      "step": 270
    },
    {
      "epoch": 0.1589945930147596,
      "grad_norm": 0.81640625,
      "learning_rate": 5.302144249512671e-05,
      "loss": 0.7339,
      "step": 272
    },
    {
      "epoch": 0.160163670904574,
      "grad_norm": 1.046875,
      "learning_rate": 5.341130604288499e-05,
      "loss": 0.8381,
      "step": 274
    },
    {
      "epoch": 0.16133274879438841,
      "grad_norm": 0.88671875,
      "learning_rate": 5.3801169590643275e-05,
      "loss": 0.774,
      "step": 276
    },
    {
      "epoch": 0.16250182668420282,
      "grad_norm": 0.796875,
      "learning_rate": 5.419103313840156e-05,
      "loss": 0.8114,
      "step": 278
    },
    {
      "epoch": 0.16367090457401726,
      "grad_norm": 0.9140625,
      "learning_rate": 5.458089668615984e-05,
      "loss": 0.7549,
      "step": 280
    },
    {
      "epoch": 0.16483998246383166,
      "grad_norm": 1.0625,
      "learning_rate": 5.4970760233918126e-05,
      "loss": 0.7393,
      "step": 282
    },
    {
      "epoch": 0.16600906035364607,
      "grad_norm": 0.88671875,
      "learning_rate": 5.536062378167641e-05,
      "loss": 0.7607,
      "step": 284
    },
    {
      "epoch": 0.16717813824346048,
      "grad_norm": 1.046875,
      "learning_rate": 5.57504873294347e-05,
      "loss": 0.8158,
      "step": 286
    },
    {
      "epoch": 0.16834721613327489,
      "grad_norm": 1.734375,
      "learning_rate": 5.6140350877192984e-05,
      "loss": 0.7403,
      "step": 288
    },
    {
      "epoch": 0.1695162940230893,
      "grad_norm": 1.1015625,
      "learning_rate": 5.653021442495127e-05,
      "loss": 0.7278,
      "step": 290
    },
    {
      "epoch": 0.1706853719129037,
      "grad_norm": 1.7578125,
      "learning_rate": 5.692007797270955e-05,
      "loss": 0.821,
      "step": 292
    },
    {
      "epoch": 0.1718544498027181,
      "grad_norm": 3.1875,
      "learning_rate": 5.7309941520467835e-05,
      "loss": 0.8276,
      "step": 294
    },
    {
      "epoch": 0.1730235276925325,
      "grad_norm": 1.1484375,
      "learning_rate": 5.769980506822612e-05,
      "loss": 0.8505,
      "step": 296
    },
    {
      "epoch": 0.17419260558234692,
      "grad_norm": 1.2265625,
      "learning_rate": 5.80896686159844e-05,
      "loss": 0.7941,
      "step": 298
    },
    {
      "epoch": 0.17536168347216133,
      "grad_norm": 1.4765625,
      "learning_rate": 5.847953216374269e-05,
      "loss": 0.8621,
      "step": 300
    },
    {
      "epoch": 0.17653076136197574,
      "grad_norm": 2.25,
      "learning_rate": 5.8869395711500976e-05,
      "loss": 0.8071,
      "step": 302
    },
    {
      "epoch": 0.17769983925179014,
      "grad_norm": 0.92578125,
      "learning_rate": 5.925925925925926e-05,
      "loss": 0.803,
      "step": 304
    },
    {
      "epoch": 0.17886891714160455,
      "grad_norm": 0.9921875,
      "learning_rate": 5.9649122807017544e-05,
      "loss": 0.7569,
      "step": 306
    },
    {
      "epoch": 0.18003799503141896,
      "grad_norm": 1.09375,
      "learning_rate": 6.003898635477583e-05,
      "loss": 0.7722,
      "step": 308
    },
    {
      "epoch": 0.18120707292123336,
      "grad_norm": 0.984375,
      "learning_rate": 6.042884990253411e-05,
      "loss": 0.7746,
      "step": 310
    },
    {
      "epoch": 0.1823761508110478,
      "grad_norm": 2.296875,
      "learning_rate": 6.0818713450292395e-05,
      "loss": 0.7292,
      "step": 312
    },
    {
      "epoch": 0.1835452287008622,
      "grad_norm": 1.140625,
      "learning_rate": 6.120857699805068e-05,
      "loss": 0.7751,
      "step": 314
    },
    {
      "epoch": 0.1847143065906766,
      "grad_norm": 0.8203125,
      "learning_rate": 6.159844054580897e-05,
      "loss": 0.7473,
      "step": 316
    },
    {
      "epoch": 0.18588338448049102,
      "grad_norm": 1.0625,
      "learning_rate": 6.198830409356725e-05,
      "loss": 0.7718,
      "step": 318
    },
    {
      "epoch": 0.18705246237030543,
      "grad_norm": 1.0390625,
      "learning_rate": 6.237816764132554e-05,
      "loss": 0.7812,
      "step": 320
    },
    {
      "epoch": 0.18822154026011984,
      "grad_norm": 0.859375,
      "learning_rate": 6.276803118908382e-05,
      "loss": 0.7552,
      "step": 322
    },
    {
      "epoch": 0.18939061814993424,
      "grad_norm": 0.765625,
      "learning_rate": 6.31578947368421e-05,
      "loss": 0.7244,
      "step": 324
    },
    {
      "epoch": 0.19055969603974865,
      "grad_norm": 1.2734375,
      "learning_rate": 6.354775828460039e-05,
      "loss": 0.8182,
      "step": 326
    },
    {
      "epoch": 0.19172877392956306,
      "grad_norm": 0.93359375,
      "learning_rate": 6.393762183235867e-05,
      "loss": 0.7578,
      "step": 328
    },
    {
      "epoch": 0.19289785181937746,
      "grad_norm": 0.875,
      "learning_rate": 6.432748538011695e-05,
      "loss": 0.8755,
      "step": 330
    },
    {
      "epoch": 0.19406692970919187,
      "grad_norm": 1.4296875,
      "learning_rate": 6.471734892787524e-05,
      "loss": 0.7436,
      "step": 332
    },
    {
      "epoch": 0.19523600759900628,
      "grad_norm": 1.859375,
      "learning_rate": 6.510721247563352e-05,
      "loss": 0.7415,
      "step": 334
    },
    {
      "epoch": 0.19640508548882069,
      "grad_norm": 1.265625,
      "learning_rate": 6.549707602339182e-05,
      "loss": 0.8082,
      "step": 336
    },
    {
      "epoch": 0.1975741633786351,
      "grad_norm": 1.4453125,
      "learning_rate": 6.58869395711501e-05,
      "loss": 0.7226,
      "step": 338
    },
    {
      "epoch": 0.1987432412684495,
      "grad_norm": 0.8203125,
      "learning_rate": 6.627680311890839e-05,
      "loss": 0.7408,
      "step": 340
    },
    {
      "epoch": 0.1999123191582639,
      "grad_norm": 1.1875,
      "learning_rate": 6.666666666666667e-05,
      "loss": 0.7752,
      "step": 342
    },
    {
      "epoch": 0.20108139704807834,
      "grad_norm": 1.0234375,
      "learning_rate": 6.705653021442495e-05,
      "loss": 0.7313,
      "step": 344
    },
    {
      "epoch": 0.20225047493789275,
      "grad_norm": 1.1015625,
      "learning_rate": 6.744639376218324e-05,
      "loss": 0.7164,
      "step": 346
    },
    {
      "epoch": 0.20341955282770716,
      "grad_norm": 0.9296875,
      "learning_rate": 6.783625730994152e-05,
      "loss": 0.7836,
      "step": 348
    },
    {
      "epoch": 0.20458863071752156,
      "grad_norm": 1.9140625,
      "learning_rate": 6.82261208576998e-05,
      "loss": 0.8936,
      "step": 350
    },
    {
      "epoch": 0.20575770860733597,
      "grad_norm": 0.8984375,
      "learning_rate": 6.861598440545809e-05,
      "loss": 0.7386,
      "step": 352
    },
    {
      "epoch": 0.20692678649715038,
      "grad_norm": 0.9609375,
      "learning_rate": 6.900584795321637e-05,
      "loss": 0.8893,
      "step": 354
    },
    {
      "epoch": 0.20809586438696479,
      "grad_norm": 0.9140625,
      "learning_rate": 6.939571150097466e-05,
      "loss": 0.715,
      "step": 356
    },
    {
      "epoch": 0.2092649422767792,
      "grad_norm": 0.94921875,
      "learning_rate": 6.978557504873294e-05,
      "loss": 0.7087,
      "step": 358
    },
    {
      "epoch": 0.2104340201665936,
      "grad_norm": 0.96484375,
      "learning_rate": 7.017543859649122e-05,
      "loss": 0.7067,
      "step": 360
    },
    {
      "epoch": 0.211603098056408,
      "grad_norm": 0.9375,
      "learning_rate": 7.05653021442495e-05,
      "loss": 0.6979,
      "step": 362
    },
    {
      "epoch": 0.2127721759462224,
      "grad_norm": 1.8046875,
      "learning_rate": 7.095516569200779e-05,
      "loss": 0.8533,
      "step": 364
    },
    {
      "epoch": 0.21394125383603682,
      "grad_norm": 0.95703125,
      "learning_rate": 7.134502923976609e-05,
      "loss": 0.6971,
      "step": 366
    },
    {
      "epoch": 0.21511033172585123,
      "grad_norm": 1.1640625,
      "learning_rate": 7.173489278752437e-05,
      "loss": 0.7152,
      "step": 368
    },
    {
      "epoch": 0.21627940961566564,
      "grad_norm": 1.3984375,
      "learning_rate": 7.212475633528265e-05,
      "loss": 0.7176,
      "step": 370
    },
    {
      "epoch": 0.21744848750548004,
      "grad_norm": 0.84375,
      "learning_rate": 7.251461988304094e-05,
      "loss": 0.7054,
      "step": 372
    },
    {
      "epoch": 0.21861756539529445,
      "grad_norm": 0.859375,
      "learning_rate": 7.290448343079922e-05,
      "loss": 0.7157,
      "step": 374
    },
    {
      "epoch": 0.21978664328510886,
      "grad_norm": 0.99609375,
      "learning_rate": 7.32943469785575e-05,
      "loss": 0.742,
      "step": 376
    },
    {
      "epoch": 0.2209557211749233,
      "grad_norm": 1.1171875,
      "learning_rate": 7.368421052631579e-05,
      "loss": 0.7016,
      "step": 378
    },
    {
      "epoch": 0.2221247990647377,
      "grad_norm": 1.3359375,
      "learning_rate": 7.407407407407407e-05,
      "loss": 0.7263,
      "step": 380
    },
    {
      "epoch": 0.2232938769545521,
      "grad_norm": 0.8515625,
      "learning_rate": 7.446393762183236e-05,
      "loss": 0.7386,
      "step": 382
    },
    {
      "epoch": 0.2244629548443665,
      "grad_norm": 1.3359375,
      "learning_rate": 7.485380116959064e-05,
      "loss": 0.7488,
      "step": 384
    },
    {
      "epoch": 0.22563203273418092,
      "grad_norm": 0.98828125,
      "learning_rate": 7.524366471734892e-05,
      "loss": 0.7304,
      "step": 386
    },
    {
      "epoch": 0.22680111062399533,
      "grad_norm": 0.90625,
      "learning_rate": 7.563352826510721e-05,
      "loss": 0.6926,
      "step": 388
    },
    {
      "epoch": 0.22797018851380974,
      "grad_norm": 1.578125,
      "learning_rate": 7.602339181286549e-05,
      "loss": 0.7258,
      "step": 390
    },
    {
      "epoch": 0.22913926640362414,
      "grad_norm": 0.82421875,
      "learning_rate": 7.641325536062378e-05,
      "loss": 0.7169,
      "step": 392
    },
    {
      "epoch": 0.23030834429343855,
      "grad_norm": 0.78125,
      "learning_rate": 7.680311890838207e-05,
      "loss": 0.6962,
      "step": 394
    },
    {
      "epoch": 0.23147742218325296,
      "grad_norm": 1.0703125,
      "learning_rate": 7.719298245614036e-05,
      "loss": 0.718,
      "step": 396
    },
    {
      "epoch": 0.23264650007306736,
      "grad_norm": 0.84375,
      "learning_rate": 7.758284600389864e-05,
      "loss": 0.7106,
      "step": 398
    },
    {
      "epoch": 0.23381557796288177,
      "grad_norm": 1.109375,
      "learning_rate": 7.797270955165692e-05,
      "loss": 0.739,
      "step": 400
    },
    {
      "epoch": 0.23498465585269618,
      "grad_norm": 0.9140625,
      "learning_rate": 7.836257309941521e-05,
      "loss": 0.7037,
      "step": 402
    },
    {
      "epoch": 0.23615373374251059,
      "grad_norm": 0.76953125,
      "learning_rate": 7.875243664717349e-05,
      "loss": 0.717,
      "step": 404
    },
    {
      "epoch": 0.237322811632325,
      "grad_norm": 1.0546875,
      "learning_rate": 7.914230019493177e-05,
      "loss": 0.7041,
      "step": 406
    },
    {
      "epoch": 0.2384918895221394,
      "grad_norm": 0.921875,
      "learning_rate": 7.953216374269006e-05,
      "loss": 0.7016,
      "step": 408
    },
    {
      "epoch": 0.23966096741195383,
      "grad_norm": 0.8515625,
      "learning_rate": 7.992202729044834e-05,
      "loss": 0.7423,
      "step": 410
    },
    {
      "epoch": 0.24083004530176824,
      "grad_norm": 0.88671875,
      "learning_rate": 8.031189083820663e-05,
      "loss": 0.7128,
      "step": 412
    },
    {
      "epoch": 0.24199912319158265,
      "grad_norm": 0.9453125,
      "learning_rate": 8.070175438596491e-05,
      "loss": 0.7144,
      "step": 414
    },
    {
      "epoch": 0.24316820108139706,
      "grad_norm": 1.1328125,
      "learning_rate": 8.109161793372319e-05,
      "loss": 0.6987,
      "step": 416
    },
    {
      "epoch": 0.24433727897121146,
      "grad_norm": 0.9296875,
      "learning_rate": 8.148148148148148e-05,
      "loss": 0.6946,
      "step": 418
    },
    {
      "epoch": 0.24550635686102587,
      "grad_norm": 0.9296875,
      "learning_rate": 8.187134502923976e-05,
      "loss": 0.6954,
      "step": 420
    },
    {
      "epoch": 0.24667543475084028,
      "grad_norm": 0.84765625,
      "learning_rate": 8.226120857699804e-05,
      "loss": 0.7507,
      "step": 422
    },
    {
      "epoch": 0.24784451264065469,
      "grad_norm": 2.171875,
      "learning_rate": 8.265107212475634e-05,
      "loss": 0.9267,
      "step": 424
    },
    {
      "epoch": 0.2490135905304691,
      "grad_norm": 0.81640625,
      "learning_rate": 8.304093567251462e-05,
      "loss": 0.7222,
      "step": 426
    },
    {
      "epoch": 0.2501826684202835,
      "grad_norm": 1.7421875,
      "learning_rate": 8.343079922027291e-05,
      "loss": 0.8919,
      "step": 428
    },
    {
      "epoch": 0.25135174631009793,
      "grad_norm": 1.140625,
      "learning_rate": 8.382066276803119e-05,
      "loss": 0.7451,
      "step": 430
    },
    {
      "epoch": 0.25252082419991234,
      "grad_norm": 0.81640625,
      "learning_rate": 8.421052631578948e-05,
      "loss": 0.7027,
      "step": 432
    },
    {
      "epoch": 0.25368990208972675,
      "grad_norm": 0.79296875,
      "learning_rate": 8.460038986354776e-05,
      "loss": 0.712,
      "step": 434
    },
    {
      "epoch": 0.25485897997954116,
      "grad_norm": 0.8984375,
      "learning_rate": 8.499025341130604e-05,
      "loss": 0.7122,
      "step": 436
    },
    {
      "epoch": 0.25602805786935556,
      "grad_norm": 0.72265625,
      "learning_rate": 8.538011695906433e-05,
      "loss": 0.6871,
      "step": 438
    },
    {
      "epoch": 0.25719713575916997,
      "grad_norm": 0.77734375,
      "learning_rate": 8.576998050682261e-05,
      "loss": 0.7278,
      "step": 440
    },
    {
      "epoch": 0.2583662136489844,
      "grad_norm": 1.1171875,
      "learning_rate": 8.61598440545809e-05,
      "loss": 0.7023,
      "step": 442
    },
    {
      "epoch": 0.2595352915387988,
      "grad_norm": 0.8359375,
      "learning_rate": 8.654970760233918e-05,
      "loss": 0.7167,
      "step": 444
    },
    {
      "epoch": 0.2607043694286132,
      "grad_norm": 0.8203125,
      "learning_rate": 8.693957115009746e-05,
      "loss": 0.7346,
      "step": 446
    },
    {
      "epoch": 0.2618734473184276,
      "grad_norm": 0.9765625,
      "learning_rate": 8.732943469785574e-05,
      "loss": 0.7223,
      "step": 448
    },
    {
      "epoch": 0.263042525208242,
      "grad_norm": 0.96875,
      "learning_rate": 8.771929824561403e-05,
      "loss": 0.7032,
      "step": 450
    },
    {
      "epoch": 0.2642116030980564,
      "grad_norm": 0.9609375,
      "learning_rate": 8.810916179337231e-05,
      "loss": 0.6804,
      "step": 452
    },
    {
      "epoch": 0.2653806809878708,
      "grad_norm": 0.875,
      "learning_rate": 8.849902534113061e-05,
      "loss": 0.7224,
      "step": 454
    },
    {
      "epoch": 0.26654975887768523,
      "grad_norm": 0.8359375,
      "learning_rate": 8.888888888888889e-05,
      "loss": 0.7138,
      "step": 456
    },
    {
      "epoch": 0.26771883676749963,
      "grad_norm": 0.859375,
      "learning_rate": 8.927875243664718e-05,
      "loss": 0.7065,
      "step": 458
    },
    {
      "epoch": 0.26888791465731404,
      "grad_norm": 0.96484375,
      "learning_rate": 8.966861598440546e-05,
      "loss": 0.7289,
      "step": 460
    },
    {
      "epoch": 0.27005699254712845,
      "grad_norm": 1.0703125,
      "learning_rate": 9.005847953216374e-05,
      "loss": 0.6917,
      "step": 462
    },
    {
      "epoch": 0.27122607043694286,
      "grad_norm": 1.0078125,
      "learning_rate": 9.044834307992203e-05,
      "loss": 0.7493,
      "step": 464
    },
    {
      "epoch": 0.27239514832675726,
      "grad_norm": 1.0546875,
      "learning_rate": 9.083820662768031e-05,
      "loss": 0.7075,
      "step": 466
    },
    {
      "epoch": 0.27356422621657167,
      "grad_norm": 0.88671875,
      "learning_rate": 9.12280701754386e-05,
      "loss": 0.6926,
      "step": 468
    },
    {
      "epoch": 0.2747333041063861,
      "grad_norm": 0.734375,
      "learning_rate": 9.161793372319688e-05,
      "loss": 0.6714,
      "step": 470
    },
    {
      "epoch": 0.2759023819962005,
      "grad_norm": 0.73046875,
      "learning_rate": 9.200779727095516e-05,
      "loss": 0.6973,
      "step": 472
    },
    {
      "epoch": 0.2770714598860149,
      "grad_norm": 0.7578125,
      "learning_rate": 9.239766081871345e-05,
      "loss": 0.7289,
      "step": 474
    },
    {
      "epoch": 0.2782405377758293,
      "grad_norm": 0.84765625,
      "learning_rate": 9.278752436647173e-05,
      "loss": 0.7296,
      "step": 476
    },
    {
      "epoch": 0.2794096156656437,
      "grad_norm": 0.99609375,
      "learning_rate": 9.317738791423001e-05,
      "loss": 0.7229,
      "step": 478
    },
    {
      "epoch": 0.2805786935554581,
      "grad_norm": 1.0859375,
      "learning_rate": 9.35672514619883e-05,
      "loss": 0.7158,
      "step": 480
    },
    {
      "epoch": 0.2817477714452725,
      "grad_norm": 0.8671875,
      "learning_rate": 9.39571150097466e-05,
      "loss": 0.6862,
      "step": 482
    },
    {
      "epoch": 0.28291684933508693,
      "grad_norm": 0.77734375,
      "learning_rate": 9.434697855750488e-05,
      "loss": 0.6535,
      "step": 484
    },
    {
      "epoch": 0.28408592722490134,
      "grad_norm": 0.90234375,
      "learning_rate": 9.473684210526316e-05,
      "loss": 0.703,
      "step": 486
    },
    {
      "epoch": 0.28525500511471574,
      "grad_norm": 0.80078125,
      "learning_rate": 9.512670565302145e-05,
      "loss": 0.7296,
      "step": 488
    },
    {
      "epoch": 0.28642408300453015,
      "grad_norm": 1.3203125,
      "learning_rate": 9.551656920077973e-05,
      "loss": 0.7061,
      "step": 490
    },
    {
      "epoch": 0.2875931608943446,
      "grad_norm": 0.76953125,
      "learning_rate": 9.590643274853801e-05,
      "loss": 0.6731,
      "step": 492
    },
    {
      "epoch": 0.288762238784159,
      "grad_norm": 1.0234375,
      "learning_rate": 9.62962962962963e-05,
      "loss": 0.7141,
      "step": 494
    },
    {
      "epoch": 0.2899313166739734,
      "grad_norm": 1.8359375,
      "learning_rate": 9.668615984405458e-05,
      "loss": 0.7874,
      "step": 496
    },
    {
      "epoch": 0.29110039456378783,
      "grad_norm": 0.97265625,
      "learning_rate": 9.707602339181286e-05,
      "loss": 0.7179,
      "step": 498
    },
    {
      "epoch": 0.29226947245360224,
      "grad_norm": 0.8984375,
      "learning_rate": 9.746588693957115e-05,
      "loss": 0.6819,
      "step": 500
    },
    {
      "epoch": 0.29343855034341665,
      "grad_norm": 0.8359375,
      "learning_rate": 9.785575048732943e-05,
      "loss": 0.688,
      "step": 502
    },
    {
      "epoch": 0.29460762823323106,
      "grad_norm": 2.21875,
      "learning_rate": 9.824561403508771e-05,
      "loss": 0.8356,
      "step": 504
    },
    {
      "epoch": 0.29577670612304546,
      "grad_norm": 2.171875,
      "learning_rate": 9.8635477582846e-05,
      "loss": 0.7733,
      "step": 506
    },
    {
      "epoch": 0.29694578401285987,
      "grad_norm": 0.8203125,
      "learning_rate": 9.902534113060428e-05,
      "loss": 0.7087,
      "step": 508
    },
    {
      "epoch": 0.2981148619026743,
      "grad_norm": 1.2109375,
      "learning_rate": 9.941520467836257e-05,
      "loss": 0.7023,
      "step": 510
    },
    {
      "epoch": 0.2992839397924887,
      "grad_norm": 1.28125,
      "learning_rate": 9.980506822612086e-05,
      "loss": 0.7087,
      "step": 512
    },
    {
      "epoch": 0.3004530176823031,
      "grad_norm": 1.328125,
      "learning_rate": 0.00010019493177387915,
      "loss": 0.7002,
      "step": 514
    },
    {
      "epoch": 0.3016220955721175,
      "grad_norm": 0.8359375,
      "learning_rate": 0.00010058479532163743,
      "loss": 0.7355,
      "step": 516
    },
    {
      "epoch": 0.3027911734619319,
      "grad_norm": 1.21875,
      "learning_rate": 0.00010097465886939573,
      "loss": 0.6983,
      "step": 518
    },
    {
      "epoch": 0.3039602513517463,
      "grad_norm": 1.65625,
      "learning_rate": 0.000101364522417154,
      "loss": 0.7016,
      "step": 520
    },
    {
      "epoch": 0.3051293292415607,
      "grad_norm": 0.765625,
      "learning_rate": 0.0001017543859649123,
      "loss": 0.7096,
      "step": 522
    },
    {
      "epoch": 0.3062984071313751,
      "grad_norm": 1.46875,
      "learning_rate": 0.00010214424951267056,
      "loss": 0.7099,
      "step": 524
    },
    {
      "epoch": 0.30746748502118953,
      "grad_norm": 1.4375,
      "learning_rate": 0.00010253411306042886,
      "loss": 0.6931,
      "step": 526
    },
    {
      "epoch": 0.30863656291100394,
      "grad_norm": 0.96875,
      "learning_rate": 0.00010292397660818713,
      "loss": 0.7427,
      "step": 528
    },
    {
      "epoch": 0.30980564080081835,
      "grad_norm": 1.234375,
      "learning_rate": 0.00010331384015594543,
      "loss": 0.6609,
      "step": 530
    },
    {
      "epoch": 0.31097471869063276,
      "grad_norm": 1.0703125,
      "learning_rate": 0.0001037037037037037,
      "loss": 0.7195,
      "step": 532
    },
    {
      "epoch": 0.31214379658044716,
      "grad_norm": 0.7109375,
      "learning_rate": 0.000104093567251462,
      "loss": 0.6735,
      "step": 534
    },
    {
      "epoch": 0.31331287447026157,
      "grad_norm": 0.82421875,
      "learning_rate": 0.00010448343079922027,
      "loss": 0.7455,
      "step": 536
    },
    {
      "epoch": 0.314481952360076,
      "grad_norm": 0.91015625,
      "learning_rate": 0.00010487329434697856,
      "loss": 0.6683,
      "step": 538
    },
    {
      "epoch": 0.3156510302498904,
      "grad_norm": 0.7265625,
      "learning_rate": 0.00010526315789473685,
      "loss": 0.6981,
      "step": 540
    },
    {
      "epoch": 0.3168201081397048,
      "grad_norm": 0.69921875,
      "learning_rate": 0.00010565302144249513,
      "loss": 0.65,
      "step": 542
    },
    {
      "epoch": 0.3179891860295192,
      "grad_norm": 2.203125,
      "learning_rate": 0.00010604288499025342,
      "loss": 0.7997,
      "step": 544
    },
    {
      "epoch": 0.3191582639193336,
      "grad_norm": 0.890625,
      "learning_rate": 0.00010643274853801171,
      "loss": 0.7143,
      "step": 546
    },
    {
      "epoch": 0.320327341809148,
      "grad_norm": 0.98046875,
      "learning_rate": 0.00010682261208576998,
      "loss": 0.6619,
      "step": 548
    },
    {
      "epoch": 0.3214964196989624,
      "grad_norm": 0.97265625,
      "learning_rate": 0.00010721247563352828,
      "loss": 0.6632,
      "step": 550
    },
    {
      "epoch": 0.32266549758877683,
      "grad_norm": 0.765625,
      "learning_rate": 0.00010760233918128655,
      "loss": 0.6844,
      "step": 552
    },
    {
      "epoch": 0.32383457547859124,
      "grad_norm": 0.828125,
      "learning_rate": 0.00010799220272904485,
      "loss": 0.698,
      "step": 554
    },
    {
      "epoch": 0.32500365336840564,
      "grad_norm": 0.94140625,
      "learning_rate": 0.00010838206627680312,
      "loss": 0.6132,
      "step": 556
    },
    {
      "epoch": 0.3261727312582201,
      "grad_norm": 0.765625,
      "learning_rate": 0.00010877192982456141,
      "loss": 0.6507,
      "step": 558
    },
    {
      "epoch": 0.3273418091480345,
      "grad_norm": 1.03125,
      "learning_rate": 0.00010916179337231968,
      "loss": 0.696,
      "step": 560
    },
    {
      "epoch": 0.3285108870378489,
      "grad_norm": 3.0,
      "learning_rate": 0.00010955165692007798,
      "loss": 0.926,
      "step": 562
    },
    {
      "epoch": 0.3296799649276633,
      "grad_norm": 1.453125,
      "learning_rate": 0.00010994152046783625,
      "loss": 0.7797,
      "step": 564
    },
    {
      "epoch": 0.33084904281747773,
      "grad_norm": 1.71875,
      "learning_rate": 0.00011033138401559455,
      "loss": 0.7379,
      "step": 566
    },
    {
      "epoch": 0.33201812070729214,
      "grad_norm": 0.9921875,
      "learning_rate": 0.00011072124756335282,
      "loss": 0.679,
      "step": 568
    },
    {
      "epoch": 0.33318719859710655,
      "grad_norm": 0.77734375,
      "learning_rate": 0.00011111111111111112,
      "loss": 0.6668,
      "step": 570
    },
    {
      "epoch": 0.33435627648692096,
      "grad_norm": 0.796875,
      "learning_rate": 0.0001115009746588694,
      "loss": 0.7378,
      "step": 572
    },
    {
      "epoch": 0.33552535437673536,
      "grad_norm": 0.8046875,
      "learning_rate": 0.0001118908382066277,
      "loss": 0.7341,
      "step": 574
    },
    {
      "epoch": 0.33669443226654977,
      "grad_norm": 0.71875,
      "learning_rate": 0.00011228070175438597,
      "loss": 0.6793,
      "step": 576
    },
    {
      "epoch": 0.3378635101563642,
      "grad_norm": 0.82421875,
      "learning_rate": 0.00011267056530214426,
      "loss": 0.6926,
      "step": 578
    },
    {
      "epoch": 0.3390325880461786,
      "grad_norm": 0.89453125,
      "learning_rate": 0.00011306042884990253,
      "loss": 0.6393,
      "step": 580
    },
    {
      "epoch": 0.340201665935993,
      "grad_norm": 0.9375,
      "learning_rate": 0.00011345029239766083,
      "loss": 0.7207,
      "step": 582
    },
    {
      "epoch": 0.3413707438258074,
      "grad_norm": 0.75390625,
      "learning_rate": 0.0001138401559454191,
      "loss": 0.6168,
      "step": 584
    },
    {
      "epoch": 0.3425398217156218,
      "grad_norm": 0.7265625,
      "learning_rate": 0.0001142300194931774,
      "loss": 0.712,
      "step": 586
    },
    {
      "epoch": 0.3437088996054362,
      "grad_norm": 0.66796875,
      "learning_rate": 0.00011461988304093567,
      "loss": 0.6532,
      "step": 588
    },
    {
      "epoch": 0.3448779774952506,
      "grad_norm": 0.734375,
      "learning_rate": 0.00011500974658869397,
      "loss": 0.6766,
      "step": 590
    },
    {
      "epoch": 0.346047055385065,
      "grad_norm": 0.72265625,
      "learning_rate": 0.00011539961013645224,
      "loss": 0.6986,
      "step": 592
    },
    {
      "epoch": 0.34721613327487943,
      "grad_norm": 0.7109375,
      "learning_rate": 0.00011578947368421053,
      "loss": 0.6867,
      "step": 594
    },
    {
      "epoch": 0.34838521116469384,
      "grad_norm": 1.484375,
      "learning_rate": 0.0001161793372319688,
      "loss": 0.8175,
      "step": 596
    },
    {
      "epoch": 0.34955428905450825,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0001165692007797271,
      "loss": 0.8122,
      "step": 598
    },
    {
      "epoch": 0.35072336694432266,
      "grad_norm": 0.96875,
      "learning_rate": 0.00011695906432748539,
      "loss": 0.7642,
      "step": 600
    },
    {
      "epoch": 0.35189244483413706,
      "grad_norm": 0.91796875,
      "learning_rate": 0.00011734892787524367,
      "loss": 0.703,
      "step": 602
    },
    {
      "epoch": 0.35306152272395147,
      "grad_norm": 0.7890625,
      "learning_rate": 0.00011773879142300195,
      "loss": 0.7031,
      "step": 604
    },
    {
      "epoch": 0.3542306006137659,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00011812865497076025,
      "loss": 0.686,
      "step": 606
    },
    {
      "epoch": 0.3553996785035803,
      "grad_norm": 0.84375,
      "learning_rate": 0.00011851851851851852,
      "loss": 0.6708,
      "step": 608
    },
    {
      "epoch": 0.3565687563933947,
      "grad_norm": 0.8515625,
      "learning_rate": 0.00011890838206627682,
      "loss": 0.8084,
      "step": 610
    },
    {
      "epoch": 0.3577378342832091,
      "grad_norm": 0.75,
      "learning_rate": 0.00011929824561403509,
      "loss": 0.6705,
      "step": 612
    },
    {
      "epoch": 0.3589069121730235,
      "grad_norm": 0.6875,
      "learning_rate": 0.00011968810916179338,
      "loss": 0.6409,
      "step": 614
    },
    {
      "epoch": 0.3600759900628379,
      "grad_norm": 0.9375,
      "learning_rate": 0.00012007797270955165,
      "loss": 0.6805,
      "step": 616
    },
    {
      "epoch": 0.3612450679526523,
      "grad_norm": 1.59375,
      "learning_rate": 0.00012046783625730995,
      "loss": 0.7129,
      "step": 618
    },
    {
      "epoch": 0.36241414584246673,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00012085769980506822,
      "loss": 0.6819,
      "step": 620
    },
    {
      "epoch": 0.36358322373228114,
      "grad_norm": 0.828125,
      "learning_rate": 0.00012124756335282652,
      "loss": 0.6424,
      "step": 622
    },
    {
      "epoch": 0.3647523016220956,
      "grad_norm": 0.7890625,
      "learning_rate": 0.00012163742690058479,
      "loss": 0.704,
      "step": 624
    },
    {
      "epoch": 0.36592137951191,
      "grad_norm": 1.3359375,
      "learning_rate": 0.00012202729044834309,
      "loss": 0.8172,
      "step": 626
    },
    {
      "epoch": 0.3670904574017244,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00012241715399610137,
      "loss": 0.6486,
      "step": 628
    },
    {
      "epoch": 0.3682595352915388,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00012280701754385965,
      "loss": 0.6772,
      "step": 630
    },
    {
      "epoch": 0.3694286131813532,
      "grad_norm": 1.1171875,
      "learning_rate": 0.00012319688109161794,
      "loss": 0.6997,
      "step": 632
    },
    {
      "epoch": 0.37059769107116763,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00012358674463937622,
      "loss": 0.6984,
      "step": 634
    },
    {
      "epoch": 0.37176676896098204,
      "grad_norm": 0.72265625,
      "learning_rate": 0.0001239766081871345,
      "loss": 0.6796,
      "step": 636
    },
    {
      "epoch": 0.37293584685079645,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0001243664717348928,
      "loss": 0.8194,
      "step": 638
    },
    {
      "epoch": 0.37410492474061086,
      "grad_norm": 0.84375,
      "learning_rate": 0.00012475633528265107,
      "loss": 0.7189,
      "step": 640
    },
    {
      "epoch": 0.37527400263042526,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00012514619883040936,
      "loss": 0.6934,
      "step": 642
    },
    {
      "epoch": 0.37644308052023967,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00012553606237816764,
      "loss": 0.6885,
      "step": 644
    },
    {
      "epoch": 0.3776121584100541,
      "grad_norm": 0.96875,
      "learning_rate": 0.00012592592592592592,
      "loss": 0.671,
      "step": 646
    },
    {
      "epoch": 0.3787812362998685,
      "grad_norm": 0.74609375,
      "learning_rate": 0.0001263157894736842,
      "loss": 0.6515,
      "step": 648
    },
    {
      "epoch": 0.3799503141896829,
      "grad_norm": 0.765625,
      "learning_rate": 0.00012670565302144252,
      "loss": 0.6833,
      "step": 650
    },
    {
      "epoch": 0.3811193920794973,
      "grad_norm": 0.76953125,
      "learning_rate": 0.00012709551656920077,
      "loss": 0.6363,
      "step": 652
    },
    {
      "epoch": 0.3822884699693117,
      "grad_norm": 0.70703125,
      "learning_rate": 0.00012748538011695908,
      "loss": 0.6848,
      "step": 654
    },
    {
      "epoch": 0.3834575478591261,
      "grad_norm": 0.93359375,
      "learning_rate": 0.00012787524366471734,
      "loss": 0.7383,
      "step": 656
    },
    {
      "epoch": 0.3846266257489405,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00012826510721247565,
      "loss": 0.7057,
      "step": 658
    },
    {
      "epoch": 0.3857957036387549,
      "grad_norm": 0.734375,
      "learning_rate": 0.0001286549707602339,
      "loss": 0.7039,
      "step": 660
    },
    {
      "epoch": 0.38696478152856933,
      "grad_norm": 0.86328125,
      "learning_rate": 0.00012904483430799222,
      "loss": 0.727,
      "step": 662
    },
    {
      "epoch": 0.38813385941838374,
      "grad_norm": 1.2265625,
      "learning_rate": 0.00012943469785575048,
      "loss": 0.7128,
      "step": 664
    },
    {
      "epoch": 0.38930293730819815,
      "grad_norm": 0.671875,
      "learning_rate": 0.0001298245614035088,
      "loss": 0.6446,
      "step": 666
    },
    {
      "epoch": 0.39047201519801256,
      "grad_norm": 0.65234375,
      "learning_rate": 0.00013021442495126704,
      "loss": 0.6434,
      "step": 668
    },
    {
      "epoch": 0.39164109308782696,
      "grad_norm": 0.671875,
      "learning_rate": 0.00013060428849902535,
      "loss": 0.642,
      "step": 670
    },
    {
      "epoch": 0.39281017097764137,
      "grad_norm": 0.9609375,
      "learning_rate": 0.00013099415204678364,
      "loss": 0.6805,
      "step": 672
    },
    {
      "epoch": 0.3939792488674558,
      "grad_norm": 0.83984375,
      "learning_rate": 0.00013138401559454192,
      "loss": 0.6923,
      "step": 674
    },
    {
      "epoch": 0.3951483267572702,
      "grad_norm": 0.921875,
      "learning_rate": 0.0001317738791423002,
      "loss": 0.7944,
      "step": 676
    },
    {
      "epoch": 0.3963174046470846,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0001321637426900585,
      "loss": 0.6876,
      "step": 678
    },
    {
      "epoch": 0.397486482536899,
      "grad_norm": 0.73046875,
      "learning_rate": 0.00013255360623781677,
      "loss": 0.6857,
      "step": 680
    },
    {
      "epoch": 0.3986555604267134,
      "grad_norm": 0.83984375,
      "learning_rate": 0.00013294346978557506,
      "loss": 0.6796,
      "step": 682
    },
    {
      "epoch": 0.3998246383165278,
      "grad_norm": 0.65234375,
      "learning_rate": 0.00013333333333333334,
      "loss": 0.6604,
      "step": 684
    },
    {
      "epoch": 0.4009937162063422,
      "grad_norm": 0.703125,
      "learning_rate": 0.00013372319688109162,
      "loss": 0.6999,
      "step": 686
    },
    {
      "epoch": 0.4021627940961567,
      "grad_norm": 1.0234375,
      "learning_rate": 0.0001341130604288499,
      "loss": 0.663,
      "step": 688
    },
    {
      "epoch": 0.4033318719859711,
      "grad_norm": 0.91796875,
      "learning_rate": 0.0001345029239766082,
      "loss": 0.6553,
      "step": 690
    },
    {
      "epoch": 0.4045009498757855,
      "grad_norm": 0.77734375,
      "learning_rate": 0.00013489278752436647,
      "loss": 0.6892,
      "step": 692
    },
    {
      "epoch": 0.4056700277655999,
      "grad_norm": 0.609375,
      "learning_rate": 0.00013528265107212476,
      "loss": 0.6529,
      "step": 694
    },
    {
      "epoch": 0.4068391056554143,
      "grad_norm": 0.7109375,
      "learning_rate": 0.00013567251461988304,
      "loss": 0.6728,
      "step": 696
    },
    {
      "epoch": 0.4080081835452287,
      "grad_norm": 0.671875,
      "learning_rate": 0.00013606237816764133,
      "loss": 0.6847,
      "step": 698
    },
    {
      "epoch": 0.4091772614350431,
      "grad_norm": 0.63671875,
      "learning_rate": 0.0001364522417153996,
      "loss": 0.7475,
      "step": 700
    },
    {
      "epoch": 0.41034633932485753,
      "grad_norm": 0.76171875,
      "learning_rate": 0.0001368421052631579,
      "loss": 0.6805,
      "step": 702
    },
    {
      "epoch": 0.41151541721467194,
      "grad_norm": 0.875,
      "learning_rate": 0.00013723196881091618,
      "loss": 0.8327,
      "step": 704
    },
    {
      "epoch": 0.41268449510448635,
      "grad_norm": 0.94921875,
      "learning_rate": 0.0001376218323586745,
      "loss": 0.663,
      "step": 706
    },
    {
      "epoch": 0.41385357299430076,
      "grad_norm": 0.6640625,
      "learning_rate": 0.00013801169590643274,
      "loss": 0.6675,
      "step": 708
    },
    {
      "epoch": 0.41502265088411516,
      "grad_norm": 0.95703125,
      "learning_rate": 0.00013840155945419105,
      "loss": 0.7458,
      "step": 710
    },
    {
      "epoch": 0.41619172877392957,
      "grad_norm": 0.66015625,
      "learning_rate": 0.0001387914230019493,
      "loss": 0.6385,
      "step": 712
    },
    {
      "epoch": 0.417360806663744,
      "grad_norm": 0.99609375,
      "learning_rate": 0.00013918128654970762,
      "loss": 0.7979,
      "step": 714
    },
    {
      "epoch": 0.4185298845535584,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00013957115009746588,
      "loss": 0.6949,
      "step": 716
    },
    {
      "epoch": 0.4196989624433728,
      "grad_norm": 0.6953125,
      "learning_rate": 0.0001399610136452242,
      "loss": 0.6713,
      "step": 718
    },
    {
      "epoch": 0.4208680403331872,
      "grad_norm": 0.64453125,
      "learning_rate": 0.00014035087719298245,
      "loss": 0.6503,
      "step": 720
    },
    {
      "epoch": 0.4220371182230016,
      "grad_norm": 0.6640625,
      "learning_rate": 0.00014074074074074076,
      "loss": 0.6641,
      "step": 722
    },
    {
      "epoch": 0.423206196112816,
      "grad_norm": 0.6640625,
      "learning_rate": 0.000141130604288499,
      "loss": 0.7159,
      "step": 724
    },
    {
      "epoch": 0.4243752740026304,
      "grad_norm": 0.69140625,
      "learning_rate": 0.00014152046783625732,
      "loss": 0.786,
      "step": 726
    },
    {
      "epoch": 0.4255443518924448,
      "grad_norm": 0.92578125,
      "learning_rate": 0.00014191033138401558,
      "loss": 0.6511,
      "step": 728
    },
    {
      "epoch": 0.42671342978225923,
      "grad_norm": 0.890625,
      "learning_rate": 0.0001423001949317739,
      "loss": 0.8375,
      "step": 730
    },
    {
      "epoch": 0.42788250767207364,
      "grad_norm": 1.609375,
      "learning_rate": 0.00014269005847953217,
      "loss": 0.7475,
      "step": 732
    },
    {
      "epoch": 0.42905158556188805,
      "grad_norm": 1.3046875,
      "learning_rate": 0.00014307992202729046,
      "loss": 0.6989,
      "step": 734
    },
    {
      "epoch": 0.43022066345170246,
      "grad_norm": 0.9453125,
      "learning_rate": 0.00014346978557504874,
      "loss": 0.6792,
      "step": 736
    },
    {
      "epoch": 0.43138974134151686,
      "grad_norm": 0.734375,
      "learning_rate": 0.00014385964912280703,
      "loss": 0.7084,
      "step": 738
    },
    {
      "epoch": 0.43255881923133127,
      "grad_norm": 0.83203125,
      "learning_rate": 0.0001442495126705653,
      "loss": 0.7263,
      "step": 740
    },
    {
      "epoch": 0.4337278971211457,
      "grad_norm": 0.78515625,
      "learning_rate": 0.0001446393762183236,
      "loss": 0.706,
      "step": 742
    },
    {
      "epoch": 0.4348969750109601,
      "grad_norm": 0.671875,
      "learning_rate": 0.00014502923976608188,
      "loss": 0.6989,
      "step": 744
    },
    {
      "epoch": 0.4360660529007745,
      "grad_norm": 0.77734375,
      "learning_rate": 0.00014541910331384016,
      "loss": 0.7097,
      "step": 746
    },
    {
      "epoch": 0.4372351307905889,
      "grad_norm": 0.58984375,
      "learning_rate": 0.00014580896686159844,
      "loss": 0.6767,
      "step": 748
    },
    {
      "epoch": 0.4384042086804033,
      "grad_norm": 0.6796875,
      "learning_rate": 0.00014619883040935673,
      "loss": 0.7045,
      "step": 750
    },
    {
      "epoch": 0.4395732865702177,
      "grad_norm": 0.828125,
      "learning_rate": 0.000146588693957115,
      "loss": 0.6492,
      "step": 752
    },
    {
      "epoch": 0.4407423644600322,
      "grad_norm": 0.609375,
      "learning_rate": 0.0001469785575048733,
      "loss": 0.6776,
      "step": 754
    },
    {
      "epoch": 0.4419114423498466,
      "grad_norm": 0.77734375,
      "learning_rate": 0.00014736842105263158,
      "loss": 0.8008,
      "step": 756
    },
    {
      "epoch": 0.443080520239661,
      "grad_norm": 0.76953125,
      "learning_rate": 0.00014775828460038986,
      "loss": 0.6677,
      "step": 758
    },
    {
      "epoch": 0.4442495981294754,
      "grad_norm": 0.6796875,
      "learning_rate": 0.00014814814814814815,
      "loss": 0.6799,
      "step": 760
    },
    {
      "epoch": 0.4454186760192898,
      "grad_norm": 0.796875,
      "learning_rate": 0.00014853801169590643,
      "loss": 0.6626,
      "step": 762
    },
    {
      "epoch": 0.4465877539091042,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0001489278752436647,
      "loss": 0.6814,
      "step": 764
    },
    {
      "epoch": 0.4477568317989186,
      "grad_norm": 0.671875,
      "learning_rate": 0.00014931773879142302,
      "loss": 0.7181,
      "step": 766
    },
    {
      "epoch": 0.448925909688733,
      "grad_norm": 0.8046875,
      "learning_rate": 0.00014970760233918128,
      "loss": 0.681,
      "step": 768
    },
    {
      "epoch": 0.45009498757854743,
      "grad_norm": 0.734375,
      "learning_rate": 0.0001500974658869396,
      "loss": 0.6625,
      "step": 770
    },
    {
      "epoch": 0.45126406546836184,
      "grad_norm": 0.69140625,
      "learning_rate": 0.00015048732943469785,
      "loss": 0.6558,
      "step": 772
    },
    {
      "epoch": 0.45243314335817625,
      "grad_norm": 0.66796875,
      "learning_rate": 0.00015087719298245616,
      "loss": 0.6968,
      "step": 774
    },
    {
      "epoch": 0.45360222124799066,
      "grad_norm": 1.171875,
      "learning_rate": 0.00015126705653021442,
      "loss": 0.7073,
      "step": 776
    },
    {
      "epoch": 0.45477129913780506,
      "grad_norm": 0.6875,
      "learning_rate": 0.00015165692007797273,
      "loss": 0.7151,
      "step": 778
    },
    {
      "epoch": 0.45594037702761947,
      "grad_norm": 0.66796875,
      "learning_rate": 0.00015204678362573098,
      "loss": 0.7192,
      "step": 780
    },
    {
      "epoch": 0.4571094549174339,
      "grad_norm": 0.70703125,
      "learning_rate": 0.0001524366471734893,
      "loss": 0.6937,
      "step": 782
    },
    {
      "epoch": 0.4582785328072483,
      "grad_norm": 0.76171875,
      "learning_rate": 0.00015282651072124755,
      "loss": 0.6859,
      "step": 784
    },
    {
      "epoch": 0.4594476106970627,
      "grad_norm": 0.7265625,
      "learning_rate": 0.00015321637426900586,
      "loss": 0.6491,
      "step": 786
    },
    {
      "epoch": 0.4606166885868771,
      "grad_norm": 0.6953125,
      "learning_rate": 0.00015360623781676414,
      "loss": 0.719,
      "step": 788
    },
    {
      "epoch": 0.4617857664766915,
      "grad_norm": 0.65625,
      "learning_rate": 0.00015399610136452243,
      "loss": 0.6906,
      "step": 790
    },
    {
      "epoch": 0.4629548443665059,
      "grad_norm": 0.72265625,
      "learning_rate": 0.0001543859649122807,
      "loss": 0.6957,
      "step": 792
    },
    {
      "epoch": 0.4641239222563203,
      "grad_norm": 0.59375,
      "learning_rate": 0.000154775828460039,
      "loss": 0.6863,
      "step": 794
    },
    {
      "epoch": 0.4652930001461347,
      "grad_norm": 0.69921875,
      "learning_rate": 0.00015516569200779728,
      "loss": 0.6845,
      "step": 796
    },
    {
      "epoch": 0.46646207803594913,
      "grad_norm": 0.625,
      "learning_rate": 0.00015555555555555556,
      "loss": 0.6393,
      "step": 798
    },
    {
      "epoch": 0.46763115592576354,
      "grad_norm": 0.67578125,
      "learning_rate": 0.00015594541910331385,
      "loss": 0.6739,
      "step": 800
    },
    {
      "epoch": 0.46880023381557795,
      "grad_norm": 0.62890625,
      "learning_rate": 0.00015633528265107213,
      "loss": 0.6919,
      "step": 802
    },
    {
      "epoch": 0.46996931170539236,
      "grad_norm": 0.6171875,
      "learning_rate": 0.00015672514619883041,
      "loss": 0.6885,
      "step": 804
    },
    {
      "epoch": 0.47113838959520676,
      "grad_norm": 0.64453125,
      "learning_rate": 0.0001571150097465887,
      "loss": 0.7075,
      "step": 806
    },
    {
      "epoch": 0.47230746748502117,
      "grad_norm": 0.6484375,
      "learning_rate": 0.00015750487329434698,
      "loss": 0.6716,
      "step": 808
    },
    {
      "epoch": 0.4734765453748356,
      "grad_norm": 0.5859375,
      "learning_rate": 0.00015789473684210527,
      "loss": 0.6595,
      "step": 810
    },
    {
      "epoch": 0.47464562326465,
      "grad_norm": 0.5390625,
      "learning_rate": 0.00015828460038986355,
      "loss": 0.6272,
      "step": 812
    },
    {
      "epoch": 0.4758147011544644,
      "grad_norm": 0.65234375,
      "learning_rate": 0.00015867446393762183,
      "loss": 0.6849,
      "step": 814
    },
    {
      "epoch": 0.4769837790442788,
      "grad_norm": 0.6953125,
      "learning_rate": 0.00015906432748538012,
      "loss": 0.6523,
      "step": 816
    },
    {
      "epoch": 0.4781528569340932,
      "grad_norm": 1.40625,
      "learning_rate": 0.0001594541910331384,
      "loss": 0.7112,
      "step": 818
    },
    {
      "epoch": 0.47932193482390767,
      "grad_norm": 0.625,
      "learning_rate": 0.00015984405458089668,
      "loss": 0.7073,
      "step": 820
    },
    {
      "epoch": 0.4804910127137221,
      "grad_norm": 0.5703125,
      "learning_rate": 0.000160233918128655,
      "loss": 0.6436,
      "step": 822
    },
    {
      "epoch": 0.4816600906035365,
      "grad_norm": 0.58203125,
      "learning_rate": 0.00016062378167641325,
      "loss": 0.669,
      "step": 824
    },
    {
      "epoch": 0.4828291684933509,
      "grad_norm": 0.94921875,
      "learning_rate": 0.00016101364522417156,
      "loss": 0.8224,
      "step": 826
    },
    {
      "epoch": 0.4839982463831653,
      "grad_norm": 0.75390625,
      "learning_rate": 0.00016140350877192982,
      "loss": 0.7526,
      "step": 828
    },
    {
      "epoch": 0.4851673242729797,
      "grad_norm": 0.64453125,
      "learning_rate": 0.00016179337231968813,
      "loss": 0.6986,
      "step": 830
    },
    {
      "epoch": 0.4863364021627941,
      "grad_norm": 0.625,
      "learning_rate": 0.00016218323586744639,
      "loss": 0.6496,
      "step": 832
    },
    {
      "epoch": 0.4875054800526085,
      "grad_norm": 0.6953125,
      "learning_rate": 0.0001625730994152047,
      "loss": 0.7165,
      "step": 834
    },
    {
      "epoch": 0.4886745579424229,
      "grad_norm": 0.58203125,
      "learning_rate": 0.00016296296296296295,
      "loss": 0.6453,
      "step": 836
    },
    {
      "epoch": 0.48984363583223733,
      "grad_norm": 0.61328125,
      "learning_rate": 0.00016335282651072126,
      "loss": 0.617,
      "step": 838
    },
    {
      "epoch": 0.49101271372205174,
      "grad_norm": 0.60546875,
      "learning_rate": 0.00016374269005847952,
      "loss": 0.6841,
      "step": 840
    },
    {
      "epoch": 0.49218179161186615,
      "grad_norm": 0.671875,
      "learning_rate": 0.00016413255360623783,
      "loss": 0.6668,
      "step": 842
    },
    {
      "epoch": 0.49335086950168056,
      "grad_norm": 0.71875,
      "learning_rate": 0.0001645224171539961,
      "loss": 0.7092,
      "step": 844
    },
    {
      "epoch": 0.49451994739149496,
      "grad_norm": 0.71484375,
      "learning_rate": 0.0001649122807017544,
      "loss": 0.6479,
      "step": 846
    },
    {
      "epoch": 0.49568902528130937,
      "grad_norm": 0.81640625,
      "learning_rate": 0.00016530214424951268,
      "loss": 0.6687,
      "step": 848
    },
    {
      "epoch": 0.4968581031711238,
      "grad_norm": 0.70703125,
      "learning_rate": 0.00016569200779727097,
      "loss": 0.6704,
      "step": 850
    },
    {
      "epoch": 0.4980271810609382,
      "grad_norm": 0.65234375,
      "learning_rate": 0.00016608187134502925,
      "loss": 0.6678,
      "step": 852
    },
    {
      "epoch": 0.4991962589507526,
      "grad_norm": 0.69140625,
      "learning_rate": 0.00016647173489278753,
      "loss": 0.6744,
      "step": 854
    },
    {
      "epoch": 0.500365336840567,
      "grad_norm": 0.79296875,
      "learning_rate": 0.00016686159844054582,
      "loss": 0.6565,
      "step": 856
    },
    {
      "epoch": 0.5015344147303814,
      "grad_norm": 0.62109375,
      "learning_rate": 0.0001672514619883041,
      "loss": 0.6499,
      "step": 858
    },
    {
      "epoch": 0.5027034926201959,
      "grad_norm": 0.84375,
      "learning_rate": 0.00016764132553606238,
      "loss": 0.6619,
      "step": 860
    },
    {
      "epoch": 0.5038725705100102,
      "grad_norm": 1.703125,
      "learning_rate": 0.00016803118908382067,
      "loss": 0.8945,
      "step": 862
    },
    {
      "epoch": 0.5050416483998247,
      "grad_norm": 0.9296875,
      "learning_rate": 0.00016842105263157895,
      "loss": 0.7597,
      "step": 864
    },
    {
      "epoch": 0.506210726289639,
      "grad_norm": 0.80078125,
      "learning_rate": 0.00016881091617933723,
      "loss": 0.6698,
      "step": 866
    },
    {
      "epoch": 0.5073798041794535,
      "grad_norm": 1.1953125,
      "learning_rate": 0.00016920077972709552,
      "loss": 0.7135,
      "step": 868
    },
    {
      "epoch": 0.5085488820692678,
      "grad_norm": 1.0859375,
      "learning_rate": 0.0001695906432748538,
      "loss": 0.8298,
      "step": 870
    },
    {
      "epoch": 0.5097179599590823,
      "grad_norm": 0.75390625,
      "learning_rate": 0.00016998050682261209,
      "loss": 0.6749,
      "step": 872
    },
    {
      "epoch": 0.5108870378488967,
      "grad_norm": 1.09375,
      "learning_rate": 0.00017037037037037037,
      "loss": 0.8318,
      "step": 874
    },
    {
      "epoch": 0.5120561157387111,
      "grad_norm": 0.7109375,
      "learning_rate": 0.00017076023391812865,
      "loss": 0.7329,
      "step": 876
    },
    {
      "epoch": 0.5132251936285255,
      "grad_norm": 0.87109375,
      "learning_rate": 0.00017115009746588694,
      "loss": 0.7127,
      "step": 878
    },
    {
      "epoch": 0.5143942715183399,
      "grad_norm": 9.375,
      "learning_rate": 0.00017153996101364522,
      "loss": 0.7969,
      "step": 880
    },
    {
      "epoch": 0.5155633494081543,
      "grad_norm": 2.015625,
      "learning_rate": 0.00017192982456140353,
      "loss": 0.8313,
      "step": 882
    },
    {
      "epoch": 0.5167324272979688,
      "grad_norm": 0.71875,
      "learning_rate": 0.0001723196881091618,
      "loss": 0.7693,
      "step": 884
    },
    {
      "epoch": 0.5179015051877831,
      "grad_norm": 0.625,
      "learning_rate": 0.0001727095516569201,
      "loss": 0.7127,
      "step": 886
    },
    {
      "epoch": 0.5190705830775976,
      "grad_norm": 0.72265625,
      "learning_rate": 0.00017309941520467836,
      "loss": 0.7283,
      "step": 888
    },
    {
      "epoch": 0.5202396609674119,
      "grad_norm": 0.70703125,
      "learning_rate": 0.00017348927875243667,
      "loss": 0.6591,
      "step": 890
    },
    {
      "epoch": 0.5214087388572264,
      "grad_norm": 0.66015625,
      "learning_rate": 0.00017387914230019492,
      "loss": 0.6794,
      "step": 892
    },
    {
      "epoch": 0.5225778167470407,
      "grad_norm": 0.57421875,
      "learning_rate": 0.00017426900584795323,
      "loss": 0.6299,
      "step": 894
    },
    {
      "epoch": 0.5237468946368552,
      "grad_norm": 0.92578125,
      "learning_rate": 0.0001746588693957115,
      "loss": 0.73,
      "step": 896
    },
    {
      "epoch": 0.5249159725266695,
      "grad_norm": 0.65625,
      "learning_rate": 0.0001750487329434698,
      "loss": 0.6538,
      "step": 898
    },
    {
      "epoch": 0.526085050416484,
      "grad_norm": 0.60546875,
      "learning_rate": 0.00017543859649122806,
      "loss": 0.7013,
      "step": 900
    },
    {
      "epoch": 0.5272541283062984,
      "grad_norm": 0.66015625,
      "learning_rate": 0.00017582846003898637,
      "loss": 0.6633,
      "step": 902
    },
    {
      "epoch": 0.5284232061961128,
      "grad_norm": 0.5234375,
      "learning_rate": 0.00017621832358674462,
      "loss": 0.6638,
      "step": 904
    },
    {
      "epoch": 0.5295922840859272,
      "grad_norm": 0.65234375,
      "learning_rate": 0.00017660818713450294,
      "loss": 0.6616,
      "step": 906
    },
    {
      "epoch": 0.5307613619757416,
      "grad_norm": 0.765625,
      "learning_rate": 0.00017699805068226122,
      "loss": 0.6935,
      "step": 908
    },
    {
      "epoch": 0.531930439865556,
      "grad_norm": 0.65625,
      "learning_rate": 0.0001773879142300195,
      "loss": 0.6795,
      "step": 910
    },
    {
      "epoch": 0.5330995177553705,
      "grad_norm": 0.87890625,
      "learning_rate": 0.00017777777777777779,
      "loss": 0.6797,
      "step": 912
    },
    {
      "epoch": 0.5342685956451848,
      "grad_norm": 0.75,
      "learning_rate": 0.00017816764132553607,
      "loss": 0.6583,
      "step": 914
    },
    {
      "epoch": 0.5354376735349993,
      "grad_norm": 0.640625,
      "learning_rate": 0.00017855750487329435,
      "loss": 0.6592,
      "step": 916
    },
    {
      "epoch": 0.5366067514248137,
      "grad_norm": 0.6328125,
      "learning_rate": 0.00017894736842105264,
      "loss": 0.7534,
      "step": 918
    },
    {
      "epoch": 0.5377758293146281,
      "grad_norm": 0.66796875,
      "learning_rate": 0.00017933723196881092,
      "loss": 0.699,
      "step": 920
    },
    {
      "epoch": 0.5389449072044425,
      "grad_norm": 0.7109375,
      "learning_rate": 0.0001797270955165692,
      "loss": 0.6759,
      "step": 922
    },
    {
      "epoch": 0.5401139850942569,
      "grad_norm": 0.58984375,
      "learning_rate": 0.0001801169590643275,
      "loss": 0.6676,
      "step": 924
    },
    {
      "epoch": 0.5412830629840714,
      "grad_norm": 0.6328125,
      "learning_rate": 0.00018050682261208577,
      "loss": 0.6734,
      "step": 926
    },
    {
      "epoch": 0.5424521408738857,
      "grad_norm": 0.72265625,
      "learning_rate": 0.00018089668615984406,
      "loss": 0.6545,
      "step": 928
    },
    {
      "epoch": 0.5436212187637002,
      "grad_norm": 0.65234375,
      "learning_rate": 0.00018128654970760234,
      "loss": 0.6863,
      "step": 930
    },
    {
      "epoch": 0.5447902966535145,
      "grad_norm": 0.54296875,
      "learning_rate": 0.00018167641325536062,
      "loss": 0.6536,
      "step": 932
    },
    {
      "epoch": 0.545959374543329,
      "grad_norm": 0.77734375,
      "learning_rate": 0.0001820662768031189,
      "loss": 0.6825,
      "step": 934
    },
    {
      "epoch": 0.5471284524331433,
      "grad_norm": 0.69921875,
      "learning_rate": 0.0001824561403508772,
      "loss": 0.7075,
      "step": 936
    },
    {
      "epoch": 0.5482975303229578,
      "grad_norm": 1.265625,
      "learning_rate": 0.00018284600389863547,
      "loss": 0.8103,
      "step": 938
    },
    {
      "epoch": 0.5494666082127722,
      "grad_norm": 1.4765625,
      "learning_rate": 0.00018323586744639376,
      "loss": 0.7081,
      "step": 940
    },
    {
      "epoch": 0.5506356861025866,
      "grad_norm": 0.734375,
      "learning_rate": 0.00018362573099415207,
      "loss": 0.6647,
      "step": 942
    },
    {
      "epoch": 0.551804763992401,
      "grad_norm": 0.98828125,
      "learning_rate": 0.00018401559454191032,
      "loss": 0.7202,
      "step": 944
    },
    {
      "epoch": 0.5529738418822154,
      "grad_norm": 1.09375,
      "learning_rate": 0.00018440545808966864,
      "loss": 0.7734,
      "step": 946
    },
    {
      "epoch": 0.5541429197720298,
      "grad_norm": 0.64453125,
      "learning_rate": 0.0001847953216374269,
      "loss": 0.6381,
      "step": 948
    },
    {
      "epoch": 0.5553119976618442,
      "grad_norm": 0.6796875,
      "learning_rate": 0.0001851851851851852,
      "loss": 0.7235,
      "step": 950
    },
    {
      "epoch": 0.5564810755516586,
      "grad_norm": 0.66796875,
      "learning_rate": 0.00018557504873294346,
      "loss": 0.8296,
      "step": 952
    },
    {
      "epoch": 0.5576501534414731,
      "grad_norm": 0.65234375,
      "learning_rate": 0.00018596491228070177,
      "loss": 0.6901,
      "step": 954
    },
    {
      "epoch": 0.5588192313312874,
      "grad_norm": 0.6796875,
      "learning_rate": 0.00018635477582846003,
      "loss": 0.6628,
      "step": 956
    },
    {
      "epoch": 0.5599883092211019,
      "grad_norm": 0.671875,
      "learning_rate": 0.00018674463937621834,
      "loss": 0.6689,
      "step": 958
    },
    {
      "epoch": 0.5611573871109162,
      "grad_norm": 0.6875,
      "learning_rate": 0.0001871345029239766,
      "loss": 0.7377,
      "step": 960
    },
    {
      "epoch": 0.5623264650007307,
      "grad_norm": 0.62890625,
      "learning_rate": 0.0001875243664717349,
      "loss": 0.7217,
      "step": 962
    },
    {
      "epoch": 0.563495542890545,
      "grad_norm": 0.65625,
      "learning_rate": 0.0001879142300194932,
      "loss": 0.6776,
      "step": 964
    },
    {
      "epoch": 0.5646646207803595,
      "grad_norm": 0.74609375,
      "learning_rate": 0.00018830409356725147,
      "loss": 0.694,
      "step": 966
    },
    {
      "epoch": 0.5658336986701739,
      "grad_norm": 0.5859375,
      "learning_rate": 0.00018869395711500976,
      "loss": 0.7305,
      "step": 968
    },
    {
      "epoch": 0.5670027765599883,
      "grad_norm": 0.671875,
      "learning_rate": 0.00018908382066276804,
      "loss": 0.6659,
      "step": 970
    },
    {
      "epoch": 0.5681718544498027,
      "grad_norm": 0.7578125,
      "learning_rate": 0.00018947368421052632,
      "loss": 0.705,
      "step": 972
    },
    {
      "epoch": 0.5693409323396171,
      "grad_norm": 0.64453125,
      "learning_rate": 0.0001898635477582846,
      "loss": 0.8348,
      "step": 974
    },
    {
      "epoch": 0.5705100102294315,
      "grad_norm": 0.65625,
      "learning_rate": 0.0001902534113060429,
      "loss": 0.7383,
      "step": 976
    },
    {
      "epoch": 0.571679088119246,
      "grad_norm": 0.609375,
      "learning_rate": 0.00019064327485380117,
      "loss": 0.7039,
      "step": 978
    },
    {
      "epoch": 0.5728481660090603,
      "grad_norm": 0.69921875,
      "learning_rate": 0.00019103313840155946,
      "loss": 0.7163,
      "step": 980
    },
    {
      "epoch": 0.5740172438988748,
      "grad_norm": 0.69140625,
      "learning_rate": 0.00019142300194931774,
      "loss": 0.6648,
      "step": 982
    },
    {
      "epoch": 0.5751863217886892,
      "grad_norm": 0.54296875,
      "learning_rate": 0.00019181286549707603,
      "loss": 0.6896,
      "step": 984
    },
    {
      "epoch": 0.5763553996785036,
      "grad_norm": 0.70703125,
      "learning_rate": 0.0001922027290448343,
      "loss": 0.7127,
      "step": 986
    },
    {
      "epoch": 0.577524477568318,
      "grad_norm": 0.8828125,
      "learning_rate": 0.0001925925925925926,
      "loss": 0.6549,
      "step": 988
    },
    {
      "epoch": 0.5786935554581324,
      "grad_norm": 0.77734375,
      "learning_rate": 0.00019298245614035088,
      "loss": 0.6581,
      "step": 990
    },
    {
      "epoch": 0.5798626333479469,
      "grad_norm": 0.578125,
      "learning_rate": 0.00019337231968810916,
      "loss": 0.6802,
      "step": 992
    },
    {
      "epoch": 0.5810317112377612,
      "grad_norm": 1.03125,
      "learning_rate": 0.00019376218323586744,
      "loss": 0.6906,
      "step": 994
    },
    {
      "epoch": 0.5822007891275757,
      "grad_norm": 0.7265625,
      "learning_rate": 0.00019415204678362573,
      "loss": 0.7383,
      "step": 996
    },
    {
      "epoch": 0.58336986701739,
      "grad_norm": 0.65234375,
      "learning_rate": 0.00019454191033138404,
      "loss": 0.7141,
      "step": 998
    },
    {
      "epoch": 0.5845389449072045,
      "grad_norm": 0.58984375,
      "learning_rate": 0.0001949317738791423,
      "loss": 0.6729,
      "step": 1000
    }
  ],
  "logging_steps": 2,
  "max_steps": 5130,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 3,
  "save_steps": 1000,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 2.57623446257664e+17,
  "train_batch_size": 1,
  "trial_name": null,
  "trial_params": null
}